You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Apr 11, 2024. It is now read-only.
I'm trying to use the DuckDB driver but running into issues.
I'm running in a linux docker container.
I've tried with a single duckdb file, and I've tried with no file specified.
Both return this error in the UI:
Logs [b44c25d9-af89-4e2d-b840-64218b2c0d0c] 2022-09-28T16:32:33-07:00 DEBUG metabase.server.middleware.log POST /api/database 400 1.6 s (0 DB calls) {:message "/tmp/libduckdb_java16810484802722090474.so: Error loading shared library ld-linux-x86-64.so.2: No such file or directory (needed by /tmp/libduckdb_java16810484802722090474.so)"}
Connect to a new DuckDB database (can be just a folder path)
See error
Expected behavior
I expect Metabase to connect to the DuckDB database and allow me to query it.
Information about your Metabase Installation:
Your browser and the version: Microsoft Edge Version 106.0.1370.34 (Official build) (64-bit)
Your operating system: Windows 11 22H2
Your databases: DuckDB
Metabase version: v0.44.4
Metabase hosting environment: Local Docker
Metabase internal database: H2
Severity
This is a complete blocker on moving my Metabase-based solution from incubation to the next phase. I'm building a free solution to be used by non-tech folks as cheaply and easily as possible. I need the simplicity and performance of DuckDB and its columnar data storage so I can provide both Metabase and the database in a single performant container.
I have a number of folks who are already using the solution, but currently it's Postgres-based and it's too expensive for them to run hosted Postgres and they are not technical enough to install and manage their own Postgres server.
Additional context
Add any other context about the problem here.
The text was updated successfully, but these errors were encountered:
The problem in the Alpine image of Metabase. The DuckDB embeds the icu lib. ICU requeres the glibc pthread lib (not the musl pthread lib installed in the Alpine by default). I tried to install different versions of the glibc library to the Metabase Alpine image but there is some issues with anyway.
Until 2.32-r0 version of the glibc, there is the std::system_error in the icu std::call_once call. I this this is the same error as in microsoft/ClearScript#331.
Starting from 2.33-r0 version of glibc there is "Error relocating /lib/ld-linux-x86-64.so.2: unsupported relocation type 37" error.
So, instead of digging into all of this issues i just created a Ubuntu based image of Metabase and added DuckDB plugin. Please see Docker section of README.md
Originally opened as metabase/metabase#25789, but redirected here.
Describe the bug
Opening issue per @flamber in this thread which has more details: https://discourse.metabase.com/t/duckdb-support/20092/2
I'm trying to use the DuckDB driver but running into issues.
I'm running in a linux docker container.
I've tried with a single duckdb file, and I've tried with no file specified.
Both return this error in the UI:
Logs
[b44c25d9-af89-4e2d-b840-64218b2c0d0c] 2022-09-28T16:32:33-07:00 DEBUG metabase.server.middleware.log POST /api/database 400 1.6 s (0 DB calls) {:message "/tmp/libduckdb_java16810484802722090474.so: Error loading shared library ld-linux-x86-64.so.2: No such file or directory (needed by /tmp/libduckdb_java16810484802722090474.so)"}
To Reproduce
Steps to reproduce the behavior:
Expected behavior
I expect Metabase to connect to the DuckDB database and allow me to query it.
Information about your Metabase Installation:
Severity
This is a complete blocker on moving my Metabase-based solution from incubation to the next phase. I'm building a free solution to be used by non-tech folks as cheaply and easily as possible. I need the simplicity and performance of DuckDB and its columnar data storage so I can provide both Metabase and the database in a single performant container.
I have a number of folks who are already using the solution, but currently it's Postgres-based and it's too expensive for them to run hosted Postgres and they are not technical enough to install and manage their own Postgres server.
Additional context
Add any other context about the problem here.
The text was updated successfully, but these errors were encountered: